Exploring Open-Vocabulary Semantic Segmentation from CLIP Vision Encoder Distillation Only.
Jun ChenDeyao ZhuGuocheng QianBernard GhanemZhicheng YanChenchen ZhuFanyi XiaoSean Chang CulatanaMohamed ElhoseinyPublished in: ICCV (2023)
Keyphrases
- semantic segmentation
- street scenes
- superpixels
- conditional random fields
- weakly supervised
- label transfer
- scene classification
- computer vision
- object categories
- object class
- bit rate
- object classes
- object detection
- image processing
- vision system
- pascal voc
- object segmentation
- multiscale
- viewpoint
- video data
- long range
- object recognition
- generative model
- image quality
- image segmentation
- motion estimation